dr rate and program design ron-02 phase 2 proposal january 31, 2006

DR Rate and Program DesignRON-02 Phase 2 Proposal

January 31, 2006

Energy & Environmental Economics, Inc. / Utilipoint International, Inc. / Freeman Sullivan & Co / Heschong Mahone Group, Inc. / Lawrence Berkeley National Laboratory

pg 2 of 45

Presentation Overview

Objectives and ProcessApproach: Phase 1 Research Findings

‣ Develop Rate Design Alternatives‣ Screen 1: Technical Potential‣ Screen 2: Resource Value‣ Screen 3: Bill Impact and Free-riders‣ Screen 4: Customer Acceptance

Phase 2 Proposal‣ Work Plan & Timeline‣ Research Team‣ Research Agenda

pg 3 of 45

Objectives and Process


pg 4 of 45

Objective of Phase 2 Research

Provide DR designs ready to pilot in CA• Research

Develop efficient, implementable rate and program designs for demand response

• ConsultationProvide a structured process to evaluate rate designs – ‘design charrette’


pg 5 of 45

Starting Point and Objective Function

Emerging CA Mkt Structure:LTRA, Day-Ahead, Balancing

Alternative Mkt Structure:Day-Ahead and/or RT Spot Price

Maximize Net Benefits

Maximize Participation Design for EmergencyDesign for Operating ReservesDesign for Planning ReservesDesign for Bill ControlDesign for Risk Mitigation Efficient, Implementable Rate Designs

Screening Process“Design Charrette”

NO YES

Revise Design

Starting Point – Market Design

Starting Point – Objective Function

DR Rate and Program Evaluation Process

Potential DR Rate and Program Designs


pg 6 of 45

Screening Process

Screen 2: Resource ValueIs it worth much? Value to system and participants?

Screen 3: Bill ImpactGiven the value, what can customers save?

Screen 4: Customer AcceptanceGiven the savings and design, will customers participate?

Candidate Rate Designs

Screen 1: Technical PotentialIs this a significant opportunity?

Research team is selected to

provide experts at each step

E3, LBNL: Stakeholder

process

HMG: Building science and loads

E3, LBNL, NA: Avoided costs and

market design

E3, NA: Rate design

FSC: Customer research


pg 7 of 45

Address major stakeholder perspectives –Utilities (Procurement, Resource planning, Marketing)–CPUC, CEC (Regulatory and Policy)–Customers (Acceptance, Complexity, Bill Impact)–ISO, WECC (Operations, and Reliability)

Characteristics of the Screening Process–Transparency–Consistency–Ease of Use–Use of publicly available data and models–Consultation with stakeholders

Screening Process


pg 8 of 45

E3 Research Team

* team leader

Team

Ren Orans*

Snuller Price

C.K. Woo

Brian Horii

Jim Williams

Roles

Overall Integration

Rate & Tariff Design

CA Regulatory Context

CA Energy Markets

Energy & Environmental Economics, Inc.

Team

Bernie Neenan*

Donna Pratt

Peter Cappers

Richard Boisvert

Roles

U.S. Energy Markets

Rate & Tariff Design

Dynamic Pricing

Program Evaluation

Utilitpoint/Neenan Associates

Team

Chuck Goldman*

Galen Barbose

Katie Coughlin

Robert Van Buskirk

Roles

RTP Rate Design

ISO DR Programs

Market Penetration

Customer Response

Lawrence Berkeley National Laboratory

Energy Markets and Policy Group

Team

Doug Mahone*

Jon McHugh

Matt Tyler

Roles

Building Science

Simulations

CA BuildingStandards

Technical Potential

Heschong Mahone Group

Team

Michael Sullivan*

Grayson Heffner

Kent Van Liere

Dan Engel

Chris Ann Dickerson

Josh Bode

Roles

Consumer Research

Participation Rates

Program Marketing

Freeman Sullivan & Company

pg 9 of 45

Approach: Phase 1 Results


pg 10 of 45

• Efficient Capacity Rationing • Ideal Rate Designs• Rate Design Criteria• Illustrative List of Candidate Designs• Phase 1 Results• Evaluation Screens

Selecting & Evaluating Candidate Designs


pg 11 of 45

Efficient Capacity Rationing

Both price rationing and quantity rationing can be efficient•Price rationing

•Types: dynamic (RTP, CPP), static (TOU)•References: Boiteux, 1949; Steiner, 1957; Joskow,

1976; Crew and Kleindorfer, 1976; Hogan, 1992•mandatory RTP is the most theoretically efficient:

P = SRMC in real time, but has practical limitations

•Quantity rationing•Types: priority service (direct load control,

interruptible, curtailable), demand subscription•References: Chao and Wilson, 1987; Spulber,

1992; Woo, 1990•can achieve 90% of efficiency of mandatory RTP.


pg 12 of 45

Ideal Rate Design

With this market structure, an ideal rate design would have the following components:

• monthly customer charge - to recover costs that vary with the number of customers on the system, such as metering, billing, and customer service

• distribution facilities charge per kW of design/contract demand – to recover the costs of local distribution facilities

• location-specific, time varying firm energy charge – to recover the time and location differentiated marginal costs of generation, transmission, and high-voltage distribution

Market Structure: ‘All-in’ Spot Market


pg 13 of 45

Under this market structure, an ideal rate design would have the following components:

• monthly customer charge - to recover costs that vary with the number of customers on the system, such as metering, billing, and customer service

• distribution facilities charge per kW of design/contract demand – to recover the costs of local distribution facilities

• generation capacity charge per kW of maximum demand - to recover the forward costs of generation capacity

• location-specific, time varying energy charge – to recover the residual time and location differentiated marginal costs of generation, transmission, and high-voltage distribution

Market Structure: Forward Capacity Market

Ideal Rate Design


pg 14 of 45

Market Structure and Rate Design

What constitutes an “ideal rate design” depends on the market

structure

‘Consumer Reports’ style evaluation

™ ? ? œ ˜GOOD → POOR

Real-time Pricing (RTP)

Firm Service Level (Demand

Subscription)

Forward Capacity Market

œ ™All-in Spot Market

™ œ


pg 15 of 45

Rate Design Criteria

Economic efficiency is one of several rate design criteria (Bonbright, 1961)

• Efficient consumption• Equitable apportionment of cost of service• Avoidance of undue discrimination • Meets utility revenue requirements• Year to year revenue stability for utility• Rate stability for customer• Simple to implement, easy to understand • Wide public acceptance


pg 16 of 45

Illustrative Candidate DR Rates

SectorOriginal Rate Type New Rate Type

Target Load

Who Controls

Enabling Technology Assumed Control Method

Conditions of dispatch

Residential Tier (E-1) Tier (E-1) A/C utilityPCT or switch

utility remotely cycles A/C or sets back thermostat, no override emergency only

Residential Tier (E-1) TOU (no tiers)whole house customer optional elasticity to TOU rate all hours

Residential Tier (E-1) DSSwhole house utility optional

firm service level, no buythrough during critical hours

economic ~40 hours per year

Residential Tier (E-1) CPP-flat (E-1) A/C shared PCTthermostat setback w/ override + elasticity to CPP rate


Small Office TOU (A-10) TOU (A-10) A/C shared PCT thermostat setback w/overrideeconomic ~40 hours per year

Retail / Large Office TOU (A-10) TOU (A-10) lighting utility

dimmable ballast

utility remotely controls customer lighting emergency only

Retail / Large Office TOU (A-10) CPP-TOU (A10)

whole building shared PCT or EMS

thermostat setback w/ override + EMS + elasticity to CPP rate


Industrial TOU (E-20) I/C (E-20) process shared optionalcurtailment call + energy manager response


Industrial TOU (E-20) I/C (E-20) process utility switch utility control of customer load emergency only

Industrial TOU (E-20) DSS + RTP process shared optionalfirm service level, RTP buythrough at all hours all hours


pg 17 of 45

‘Consumer Reports’ Style Evaluation

™? ?Good Poor

Avoids free Control ValueSimplicity riders Utility Customer Planning Operating Emergency

Residential TOU Rate Res Tier w/ PCT or Switch Res Demand Subscription Res CPP Com PCT w/ Override Commercial Emergency Commercial CPP Com Demand Subscription Ind Emergency Ind CPP Ind I/C - Customer Control Ind I/C - Utility Control Real Time Pricing Dmd Subscription w/ RTP

Voluntary Demand Subscription Service with an enabling technology (PCT) was identified as best option for existing

conditions

Summary of Phase 1 Results


pg 18 of 45

Residential Demand Subscription

• Customers subscribe to a firm service level with a monthly $/kW-mo charge. Customers must not exceed this level during notice periods.

• Energy rates are reduced, but can maintain tier structure to minimize free rider bill impacts.

• Customers can reduce their bills by subscribing to a level below their maximum demand

• Subscription level can be enforced with a limiter device, or with a very high price for excess usage.

• If customers subscribe to their maximum demand, they need not alter their behavior. For those customers that subscribe to lower levels of demand, however, some education and effort will be required for them to reduce their loads.

™? ?Good Poor

Avoids free Control ValueSimplicity riders Utility Customer Planning Operating Emergency

Res Demand Subscription ? ? ™ ? ™ ™ ™

Summary of Phase 1 Results

pg 19 of 45

Screen 1: Technical Potential


pg 20 of 45

Example: Top Ten Contributors to System Peak

0

1,000

2,000

3,000

4,000

5,000

6,000

7,000

8,000

9,000

MW

at

peak

Data from CEC Demand Forecast Office

Phase 1 end-uses selected

Technical potential screen


pg 21 of 45

End Use Peak Demand (MW) PctCom AC 8,139 15%Res AC 7,917 14%Assembly Industry 6,373 11%Com Light 6,322 11%Com Misc 3,674 7%Res Misc 3,556 6%TCU Buildings 2,508 4%Ag & Water Pumping 2,487 4%Process Industry 2,289 4%Res Refrigerator 2,175 4%Com Ventilation 1,946 3%Res Cooking 1,433 3%Mining and Construction 1,095 2%Res Clothes Dryers 1,086 2%Com Refrigerators 996 2%Res Swimming Pool Pump 588 1%Res Television 548 1%Res Single Family Hot Water 409 1%Res Freezer 400 1%Res Dishwashing 377 1%Com Office Equipment 314 1%Res Spa Pump 270 0%Res Multi Family Hot Water 209 0%Res Water beds 162 0%Res Clothes washer 131 0%Com Domestic hot water 129 0%Com Exterior Lighting 111 0%Com Cooking 102 0%Res Spa Heater 49 0%Res Solar Hot Water Pump 36 0%Res Pool Heating 9 0%Res Solar Domestic Hot Water 4 0%Res Solar Pool 0 0%Total 55,846 100%

2001 California Peak Demand

Example: non-time critical customer loads

• In addition to the large end-use segments, there are smaller load segments with less time critical customer loads that have potential for demand response.• Pool pumps• Electric water heat

• ‘low-hanging’ fruit.



pg 22 of 45

Example: Residential PCTs

Building Type Technology

Comm. SF or dwell. units

Appliance Stock

W/sf or kW/unit savings

Total Peak kW

Technical Savings

multiplier

Total technical response kW/sector

Residential SectorSingle Family Residential A/C PCT 7.77E+06 2.91E+06 0.9 2.54E+06 1.95E+06

PG&E2003 A/C PCT 3.17E+06 1.03E+06 8.30E+05 6.40E+05Forecast Zone 1 7.95E+04 0.5 3.98E+04 0.69 2.74E+04Forecast Zone 2 1.62E+05 0.8 1.30E+05 0.85 1.10E+05Forecast Zone 3 3.22E+05 0.9 2.90E+05 0.85 2.47E+05Forecast Zone 4 4.08E+05 0.8 3.27E+05 0.69 2.25E+05Forecast Zone 5 5.42E+04 0.8 4.34E+04 0.69 2.99E+04

SMUD2003 A/C PCT 3.47E+05 2.51E+05 2.26E+05 1.92E+05Forecast Zone 6 3.47E+05 2.51E+05 0.9 2.26E+05 0.85 1.92E+05

More Utilities and forecast zones below

Approximately one third A/C saturation in PG&E territory

Savings per home are climate (forecast zone) dependent

Fraction of A/C on during peak is climate

dependent



pg 23 of 45

Technical potential screenExample: PCT DOE-2 results 2 – 6 pm

0

0.5

1

1.5

2

2.5

3

3.5

0 2 4 6 8 10 12 14 16 18 20 22 24

2Flr-EW HomeSame w/ PCT

Analysis Repeated for Each of the Dispatched Days

Period of curtailment

Average 1 kW savings

Rebound from PCT release


pg 24 of 45


Illustrative findings

List is illustrative, not exhaustive

SectorDefinition

End-use, or Technology

Coincident Peak kW

Total technical response kW/sector

Single Family Residential A/C PCT 2,541,319 1,945,859 Single Family Residential A/C cycle off 6,895,801 2,744,013 Single Family Residential Pool Pump 580,000 580,000Small Office A/C PCT 551,791 551,791 Small Office Lighting 414,264 207,132 Retail Lighting 1,433,955 716,978 Colleges Lighting 193,614 96,807 Industrial Sector Existing Control 6,201,000 306,022

pg 25 of 45

Screen 2: Resource Value


pg 26 of 45

Resource Value Screen

Offer to customers is driven by the value of the load reduction.

Resource value screen evaluates whether the rate design provides significant value.

Rate or Program Structure

Value of Rate or Program

Rate or Program Design

Customer Bill Change .


pg 27 of 45


Control Utility(Quantity

)

Utility w/ customer over-ride

Customer(Pricing)

Emergencyreserve

K1D1VE K4D4VE K7D7VE

Operating reserve

K2D2VO K5D5VO K8D8VO

Planning reserve

K3D3VP K6D6VP K9D9VP

Likely high value

Uncertain

Likely zero value

Components of Value

Vx = capacity value $/kW

Kx = enrolled kW

Dx = derating factor

(equivalent reliability)

Value Matrix


pg 28 of 45

Uncertainty in customer-controlled DR

• WECC has stringent standards for counting load resources towards reserve requirements• WECC Nonspinning reserve currently requires

that load be interruptible within 10 minutes.WECC Minimum Operating Reliability Standards Standards BAL-STD-001-0-WECC — Real Power Balancing Control PerformanceWRS2. Acceptable types of nonspinning reserve. The nonspinning reserve obligations identified in WR1, WRS1.1, and WRS1.2, if any, can be met by use of the following:

(a) load which can be interrupted within 10 minutes of notification

(b) interruptible exports

(c) on-demand rights from other entities or Control Area/Balancing

(d) spinning reserve in excess of requirements in WR1

(e) off-line generation which qualifies as nonspinning reserve (see definition)



pg 29 of 45

Example: Derating a PCT load reduction

operator issues curtailment request

customer AC is installed and load is present

PCT resets temperature setpoint

PCT commands AC to turn off at setpoint

customer chooses not to override

curtailment manager issues command

RF setback command sent to PCTs

AC cycles as intended, reduces load

PCT receives RF setback command

customer PCT is installed and operable

load reductions aggregated at zonal level

override indic-ator, smart meter installed

confirmation of load reduction sent to operator

operator receives confirmation

1.00 0.99 0.99 0.80

0.970.950.990.99

0.75 0.90

DERATING FACTORS ARE MULTIPLIED TOGETHER

OVERALL DERATING FACTOR = 0.48

Derating factors at each stage are in red


pg 30 of 45

˜ œ ? ? ™POOR → GOOD

Net Social Welfare

•Rates not based on market value result in some efficiency loss•Voluntary rates allow preference discovery and preference matching•*High transaction costs (e.g. technology) can lower welfare gains of mandatory rates

*


pg 31 of 45

Screen 3: Bill Impact and Free-Riders


pg 32 of 45

Bill Impact Screen

Rate or Program Structure

Value of Rate or Program

Rate or Program Design

Rate or program design affects customers bill savings.

Bill impact screen evaluates whether the rate will provide the customer significant savings.Customer Bill Change

.


pg 33 of 45

Example of Demand Subscription

Load Reduction by Reliability Level (kW) for a 30% annual load factor customersDemand subscription Monthly kWhlevel (% of annual max) 100 200 300 400 500 600 800 1000 1500 2000 2500 3000

100% - - - - - - - - - - - - 90% 0.05 0.09 0.14 0.18 0.23 0.27 0.37 0.46 0.69 0.91 1.14 1.37 80% 0.09 0.18 0.27 0.37 0.46 0.55 0.73 0.91 1.37 1.83 2.28 2.74 70% 0.14 0.27 0.41 0.55 0.69 0.82 1.10 1.37 2.06 2.74 3.43 4.11 60% 0.18 0.37 0.55 0.73 0.91 1.10 1.46 1.83 2.74 3.65 4.57 5.48 50% 0.23 0.46 0.69 0.91 1.14 1.37 1.83 2.28 3.43 4.57 5.71 6.85 40% 0.27 0.55 0.82 1.10 1.37 1.64 2.19 2.74 4.11 5.48 6.85 8.22

Bill Discount (for 30% load factor customers)Demand subscription Monthly kWhlevel (% of annual max) 100 200 300 400 500 600 800 1000 1500 2000 2500 3000

100% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0%90% 4% 4% 4% 4% 4% 3% 3% 3% 2% 2% 2% 2%80% 8% 8% 8% 8% 8% 7% 6% 5% 4% 4% 4% 3%70% 12% 12% 12% 12% 12% 10% 9% 8% 6% 6% 5% 5%60% 17% 17% 17% 16% 16% 14% 12% 10% 8% 7% 7% 7%50% 21% 21% 21% 21% 20% 17% 15% 13% 10% 9% 9% 8%40% 25% 25% 25% 25% 24% 21% 18% 15% 12% 11% 11% 10%

• Demand subscription charge = $10.92/kW-mo. (30% of total revenues). • Other residential tier $/kWh rates reduced by 30%.• Rate is revenue neutral based on average DS of 95% of max demand.• Large users receive a relative lower % discount because the DS charge in this

example is constant ($/kW-mo), while the average rate increases with size.

Bill Impact Screen


pg 34 of 45

Free Rider Potential

Mandatory

enrollment

Voluntary

enrollment

Customer control ™ œUtility control ™ ?

˜ œ ? ? ™POOR → GOOD

TOU

DSS

Voluntary enrollment in customer control DR programs (pricing) is the most subject to free-riders.

Free Rider Screen

pg 35 of 45

Screen 4: Customer Acceptance


Phase 1 research on customer acceptance (conducted by FSC):

1. Assessed factors affecting enrollment and response

2. Calculated projected enrollment rates for illustrative designs

Customer Acceptance Screen


pg 37 of 45

ResponsivenessResponsiveness

Enrollment and Response Estimation Framework

RATE/PROGRAM DESIGNSRATE/PROGRAM DESIGNS

Enrollment RateEnrollment Rate

OPT-OUTOPT-OUT

Attractiveness of DesignBill Savings, Volatility, Frequency

of Events, Tech Package



Attractiveness of Available Alternatives


Participation CostsPartial Outage, Transaction, Risk,

and Equipment costs


and Equipment costs

Enrollment RateEnrollment Rate

OPT-INOPT-IN






and Equipment costs


and Equipment costs



DEMAND RESPONSE IMPACT

DEMAND RESPONSE IMPACT

Customer FactorsAwareness, Attitude, Intention,

Behavior


Behavior


Behavior


Behavior Impacts vary over time Within given operations Across repetitive operations Across implementation cycle

Effectiveness of TargetingEffectiveness of Targeting

Enabling TechnologyEnabling Technology

Size and Volatility of Price


Reliability of ResponseReliability of Response

Burn-outBurn-outChurnChurn


pg 38 of 45

Revealed Preference Models

Revealed Preference Models

Reliability of Estimation Methods

Enrollment

Less Reliable Estimate

Enrollment

More Reliable Estimate

ResponsivenessMore Reliable Estimate

Experience in Similar Programs

Experience in Similar Programs

Recruitment into Pilots

Recruitment into Pilots

Participant Surveys

Participant Surveys

RATE/PROGRAM DESIGNS

Enrollment Rate

OPT-OUT




Participation CostsPartial Outage, Transaction, Risk, and Equipment costs

Responsiveness Enrollment Rate

OPT-IN



Participation CostsPartial Outage, Transaction, Risk, and Equipment costs

Effectiveness of Targeting


DEMAND RESPONSE IMPACTS ( MW)

Reliability of Response

Burn-outChurn

Reliability of Response

Burn-outChurn

Enabling Technology


Enabling Technology



Behavior


Behavior

ResponsivenessLess Reliable Estimate

Stated Preference Models

Stated Preference Models

Pilot/Field Testing

Pilot/Field Testing

Price Elasticities

Price Elasticities

Experience with the

Real Thing

Experience with the

Real Thing


pg 39 of 45

Steady State Enrollment Rates for Sample DR Rates/Programs

Momentum Based (Assumes 100% awareness, No Transaction

Costs, No Measurement of Intention)

Actual Program Experience

Lower Bound Upper Bound Lower Upper

Residential TOU Rate 33.0% 47.0% 6% 10%

Residential Tier w/ PCT or Switch 41.0% 47.0% 5% 15%Residential Demand Subscription 34.0% 47.0% 5% 10%Residential CPP-F 34.0% 47.0% 5% 10%

Commercial PCT w/ Override 37.0% 39.0% 3% 10%

Commercial Emergency - utility control UnknownCommercial CPP 34.0% 39.0% 5% 10%Commercial Demand Subscription 34.0% 39.0% 5% 10%

Industrial I/C - Customer Control 5% 20%

Industrial I/C - Utility Control 0% 10%Industrial Real Time Pricing 1% 3%

Rate/Program DesignSector

Illustrative ‘Opt-in’ Enrollment Estimates


pg 40 of 45

Customer Acceptance Findings

•Targeting strategies, enabling technologies, and marketing have a substantial effect on both enrollment and response rates

•Momentum’s enrollment estimates are based on customer preference modeling assuming 100% awareness.

•FSC’s lower estimates are based on actual program experience. These estimates are uncertain but the ranges are reasonable.

Phase 2 screen will build upon participation literature and experience and will be augmented by quantitative research.

pg 41 of 45

Phase 2 Proposal


pg 42 of 45

Phase 2 Proposal Summary

Deliverables:• suite of prototype DR rate and program

designs for California• set of screening tools to evaluate any DR

program designProcess:• consultative stakeholder process• E3 takes lead in framing questions,

conducting research, writing drafts


pg 43 of 45

Phase 2 General Approach

• Engage stakeholders throughout process

• Consider wide spectrum of candidate designs

• Make screens flexible to allow for future evolution of market and regulation

• Make rate design consistent with DR value

The success of Phase 2 should be measured

by implementation of DR designs

with broad stakeholder support


pg 44 of 45

Phase 2 Proposed Tasks

1. Develop MOU2. Finalize Research Plan3. Develop Screening Tools and Designs4. Initial Screening of Candidate Programs5. Complete Screening Tools and Program

Designs6. Final Screening of Prototype Programs7. Final Report


pg 45 of 45

Phase 2 Timeline

DR Rate Design Phase 2 ScheduleMonth After Award

1 2 3 4 5 6 7 8 9

Task 1Develop MOU and initiate consultative process

Task 2Review Phase 1 and finalize Phase 2 research plan

Task 3Develop screening tools and program designs

Task 4Initial screening of candidate programs

Task 5Complete screening tools and program designs

Task 6Final screening of prototype programs

Task 7Final report

Deliverable (Draft or Final)

dr rate and program design ron-02 phase 2 proposal january 31, 2006

Documents